Breaking and Seating of Rigid Pavements

Transcription

1 TRANSPORTATION RESEARCH RECORD Breaking and Seating of Rigid Pavements GARY w. SHARPE, MARK ANDERSON, AND ROBERT. DEEN* Breaking and seating have been used extensively In Kentuky to rehabilitate portland ement onrete pavements. Experiene over 3 or 4 yr with this type of design and onstrution Is summaried and reported. Breaking to a range of nominal fragments Is evaluated, and a report on the evaluation of two roller weights for seating is given. Also desribed Is the use of dynami. defletions to gauge the effetiveness of the breaking and seatmg proess and to measure the appropriateness of the asphalti onrete overlay. Rigid (portland ement onrete) pavements are deteriorating rapidly in many areas of the ountry. Spalling, raking, joint deterioration, and faulting at joints and raks are ommon and lead to deteriorating ride quality and safety as well as inreasing maintenane osts. Joint repairs or full-sale replaement result in signifiant apital expenditures and lengthy delays for travelers. Two tehniques used for rehabilitating rigid pavements are reyling and overlaying. Reyling may be done at a entral plant or may be arried out in plae. Centralied reyling typially involves pulveriing the existing onrete pavement, removing the fragmented material, proessing the material (rushing, grading, removing steel, stok piling), and using all or a portion of the material as aggregate in a new onrete or hot-mix asphalt mixture. In-plae reyling onsists of onverting the existing onrete pavement to a base and then overlaying it with either asphalti onrete or portland ement onrete. Refletion raking of existing raks and joints of the underlying pavement is a major problem when asphalti onrete overlays are used over unbroken rigid pavements. Tehniques employed speifially to redue or prevent refletion raking have not been ompletely suessful. Proedures urrently reeiving attention inlude (a) breaking and seating the existing onrete pavement followed by plaement of a relatively thik (more than 4 in.) asphalti onrete overlay and (b) plaement of a rak-relieflayer followed by a moderately thik overlay (less than 4 in.) of asphalti onrete. A typial rak-relief layer onsists of 3 to 4 in. of opengraded bituminous material plaed over an existing rigid pavement. Another 3 to 4 in. of asphalti onrete base and surfae typially are plaed over the rak-relief layer (J). In-plae reyling of rigid pavements has beome popular in Kentuky in reent years. Speifi methods have varied but generally onsist of breaking and seating the rigid pavement *Deeased. Transportation Researh Program, University of Kentuky, Lexington, Ky followed by overlaying with asphalti onrete. Nominal sies of fragments vary from 1 /2 x 3 ft to 4 x 6 ft, and overlay thiknesses used nationally range from 2 3 /4 in. to 7 3 /4 in. Pries for breaking and seating have varied from $0.25/yd2 to $2.00 or more/yd2 (1-3). Types of breaking devies inlude a pile driver with a modified shoe, a transverse drop-bar (guillotine) hammer, a whip hammer, an impat hammer, and a resonant pavement breaker. There are also many different methods of seating broken onrete partiles. Roller sies have varied from 44, lb to 100, lb (J). Pneumati-tired rollers weighing 30 to 50 tons are more ommonly used, although there has been some experimentation with vibratory rollers of the steelwheeled and sheepsfoot varieties. BREAKING AND SEATING IN KENTUCKY Kentuky has embarked on an extensive breaking and seating program to rehabilitate deteriorated portland ement onrete pavements. Between 1982 and 1986, over 750 lane-miles of pavement have been broken, seated, and overlaid with asphalti onrete. Performane has been good; as a result, the pratie ontinues routinely. Road Rater defletion measurements have been obtained for a number of pavement setions before breaking, after breaking but before seating, at various stages during seating, after seating, and periodially after overlaying. Additionally, defletion measurements have been obtained at various phases of the seating ativities for both 50-ton and 35-ton pneumati rollers. A detailed visual survey has been onduted for a number of setions. Findings of these evaluations will be summaried in this paper. These data will ontribute to evaluation of the longterm performane of these pavements and of the effetiveness of breaking and seating proedures. Additionally, these data will be helpful in the development of rational tehniques for determining overlay thikness requirements over broken and seated pavements. Currently, Kentuky thikness design determinations are based on the assumption that the broken portland ement onrete will perform in the same manner as a onventional dense-graded aggregate base. The validity of this assumption needs to be determined. Breaking Patterns The ondition of the existing rigid pavement may signifiantly influene the manner in whih a pavement will frature. The

2 TRANSPORTATION RF.SEARCH RECORD resultant breaking pattern apparently is a funtion of the energy absorbed by the slab and the way in whih the energy is dissipated throughout the slab and pavement struture. Dissipation of energy is dependent on the strength and thikness of the existing onrete, joint and rak spaing and ondition, and degree of deterioration of the slab. Other fators may inlude temperature and time of day, whih may affet the extent and degree of urling and warping that may alter resulting pavement raking patterns. For example, peuliar pavement breaking patterns (longitudinal fraturing resulting in a series of "beams") have been observed during extended periods of high temperature. High temperatures may result in exessive ompressive stresses at joints, whih then may alter pavement breaking harateristis. The appropriate nominal sie of fragmentation remains ontroversial. The sie of fragments has a diret impat on design onsiderations as well as on the long-term performane of the overlay. Small fragments will most ertainly redue and possibly eliminate refletive raking in the asphalti onrete overlay but use the least strutural potential of the existing portland ement onrete pavement. Conversely, very large fragments may maximie the strutural potential of the existing portland ement onrete but may be large enough to permit thermal movements of the existing piees and thereby maintain the potential for refletive raking. Large fragments may also have more potential for roking as a result of ineffetive seating and may therefore inrease the potential for raking of the overlay. Researh in Kentuky has involved three ranges of nominal fragment sies for raked onrete: 1. 3 to 12 in., to 24 in., and to 36 in. Current Kentuky speifiations (4) require pavements to be broken to a nominal 24-in. sie and permit up to 20 perent of the fragments to exeed 24 in. Piees larger than 30 in. are not permitted Researh is ontinuing to determine the optimum sie for fragmenting portland ement onrete pavements. No definite onlusions appear to have been reahed at this time. Experiene in Kentuky generally favors the 18- to 24-in. fragments. Current speifiations require viewing fragmentation patterns of a dry surfae (4). There is also no uniform proedure to determine whether a broken slab meets required speifiations. Two proedures have been used to evaluate the extent of breaking: 1. Visual evaluation by ounting the number of partiles and measuring the maximum dimensions of the largest partiles, and 2. Comparison of defletion measurements before and after breaking using a Road Rater. Visual evaluations are more readily adaptable to apabilities of onstrution inspetion personnel but are subjet to ontroversy beause of subjetivity. They are used routinely for aeptane or rejetion of the breaking pattern. Defletion testing has been used only for verifiation of the effetiveness of breaking and seating. Early Kentuky plan notes allowed the raking pattern to be viewed by wetting the pavement surfae. Wetting the surfae presented inspetion problems beause it is not pratial to ontinually wet the surfae for viewing the raking pattern. Some raking may be observed without the aid of a wetted surfae and is dependent on the harateristis of the unbroken slab, equipment used to break and seat, and ondition of underlying layers. Current speial provisions (4) require the broken pavement to be viewed without the aid of a wetted surfae. Defletion testing provides a more objetive and definitive omparison of before-and-after onditions. The prinipal problem assoiated with defletion testing for aeptane or rejetion is the availability of defletion testing equipment for onstrution personnel and the level of experiene and expertise required to ollet and interpret defletion measurements. Breaking Equipment Three types of pavement breakers have been used in Kentuky: (a) pile-driving hammer, (b) transverse-bar drop hammer (guillotine), and () whip hammer. The pile -driving hammer and the whip hammer typially result in longitudinal and diagonal raking, whereas the transverse-bar drop hammer typially produes transverse raking of the existing portland ement onrete pavement. The most ommon pavement breaker urrently in use in Kentuky is the modified diesel pile-driving hammer. The hammer typially is mounted in a rolling arriage and is towed by a trator. The fore or energy of impat may be hanged by throttling the flow of fuel to the hammer. The greater the fuel input to the hammer, the greater the fore applied to the pavement. Generally the firing rate for a hammer remains onstant. As suh, the number of blows applied to the pavement may be modified by varying the speed of the towing vehile. The breaking pattern is a funtion of the energy applied to the pavement slab. One method of "measuring" the energy input is to determine the total number of blows applied to the pavement at a onstant fore or impat level for the hammer. Experiene in Kentuky has shown that 18- to 24-in. fragments may be ahieved when the pile-driving hammer traverses a slab with three or four passes per lane width equally spaed transversely aross the slab and the interval between impat blows of the hammer is 12 to 18 in. The transverse spaing of passes, interval between impat blows, number of passes, and hammer throttle setting are funtions of the ondition and thikness of the existing portland ement onrete and the quality of the subgrade. The throttle setting for a piledriving hammer should be at a level suffiient to frature the pavement yet not so large as to reate punhing and deep indentations. Additional experiene in Kentuky has indiated that fragment sies of 30 to 36 in. may be ahieved with two or three passes of a pile hammer at an interval of 12 to 18 in. between impat blows. Similarly, fragments of 3 to 12 in. may result from seven to eight passes and the same 12- to 18-in. interval between impat blows. One other fator affeting the breaking pattern when using the pile-driving hammer is the shape of the head or "shoe" that strikes the pavement. Breakers used in Kentuky typially have a plate type of shoe to prevent or minimie penetration or

3 Sharpe el al. punhing into the surfae of the ex1stmg portland ement onrete pavement. The most effetive shoe is apparently a square (on the order of 18 in. square) rotated 45 degrees to the diretion of travel. This shape apparently ontributes to diagonal breaking interonneted with longitudinal raks to form the desired pattern. The whip hammer onsists of an impat hammer attahed to the end of a leaf-spring arm. The whip hammer may be moved in the horiontal as well as the vertial diretion. The impat fore is developed by the whipping ation of the leaf-spring arm and hammer head. The energy is transmitted to the pavement by a base plate or shoe in muh the same manner as that noted with the pile-driving hammer. Typially, the plate will have a diamond, square, or retangular shape. The whip hammer typially is mounted on the rear of a truk and usually is equipped with dual ontrols, permitting use by only one operator. The fore developed by the whip hammer is apparently a funtion of the pressure in the hydrauli system and the resilieny and nwnber of leaf springs supporting the hammer head. As is seen with the pile-driving hammer, the resulting raking pattern is a funtion of the total number of blows applied to the pavement. Blows from the whip hammer typially are applied in a more random manner than they are for the pile-driving hammer. This provides for greater potential of a random raking pattern but at the same time makes it more diffiult to input a onsistent level of impat energy. The whip hammer may be maneuvered in an ar, typially providing a overage of approximately an 8-ft ar. An 18- to 24-in. breaking pattern may usually be ahieved with one blow of the whip hammer per square foot of pavement surfae area. The whip hammer has not yet been used in Kentuky to break rigid pavement to other sies. As noted with the pile-driving hammer, the speifi fragment sie will vary from pavement setion to pavement setion. The transverse drop-bar (guillotine) hammer has been used to break one setion (approximately 50 lane-miles) of onrete pavement in Kentuky. The drop bar (blade) typially weighs 5 to 7 tons and the drop is usually 18 in. The operator varies the speed of travel and thereby ontrols the interval between impats. The fore of impat may be varied by hanging the height of the drop (J, 2). Seating Seating the fragments is neessary to ensure a stable foundation for the asphalti onrete overlay. With inadequate seating, individual fragments tend to rok, inreasing the potential for refletion raking. With pavement breaking, seating requirements and harateristis may vary with fragment sie, quality, and harateristis of the existing pavement and quality of the subgrade. The objetive of seating is to plae all fragments in ontat with the supporting aggregate base or subgrade. Experiene so far has indiated that the most effiient seating of a broken portland ement onrete pavement may be aomplished by rolling with a heavy pnewnati-tired roller. Typial roller sies vary from 30 to 50 tons. Steel-wheeled (stati and vibratory) rollers have been used but have not been fully effetive beause of bridging over fragments. An 8-ton steel-wheeled vibratory roller was speified for the first projet in Kentuky but this roller proved inadequate. Roller requirements were modified by a onstrution hange order to use a 30-ton pnewnati-tired roller. Subsequent projets required seating by a 50-ton pneumati-tired roller. Reent evaluations, however, have indiated that the 30-ton pneumati-tired roller is almost as effetive. Currently, a 30-ton pnewnati-tired roller is the smallest roller permitted. EVALUATIONS Effetiveness of Breaking A simplified tehnique has been used for evaluating defletions obtained before, during, and after breaking portland ement onrete pavement as well as after paving. Examples of defletions of two pavements are presented in Tables 1 and 2. The tables present average field measured defletions as well as theoretially simulated defletions and assoiated layer moduli. Field data in Tables 1 and 2 were used to determine information presented in Table 3, whih summaries ratios of defletions after breaking (but before overlaying) to defletions before breaking. The ratios also are summaried in Figure 1. There appears to be a relationship among fragment sie, effetive stiffness modulus, and ratio of defletions (after breaking to before breaking). Effetiveness of Seating Defletion measurements were obtained before breaking and after various intervals during rolling with the 30-ton roller used for the first Kentuky projet and for a 35-ton and 50-ton roller for a subsequent projet. Results of the latter evaluation are summaried in Figures 2, 3, and 4. Data from three loations (midslab, opposing third points, and opposing edges or omers) are presented The average defletions shown are for all slabs tested and for all four Road Rater sensors. Initially, average defletion urves were plotted for eah sensor, but the similarity of the urves suggested that they ould be ombined into the average urves shown. Data indiate the following general trends: 1. An inrease in defletions after initial roller passes, 2. A redution or stabiliation of defletions with additional roller passes, and 3. An inrease in defletions with a large nwnber of roller passes. At the midslab and third-point loations, the two rollers had similar average defletions, with the 35-ton roller atually giving more onsistent values. At the edges, however, the 35- ton roller did not appear to seat the broken pavement as well as did the 50-ton roller. This is not surprising, beause the 35-ton roller was not as wide as the 50-ton roller. In the omparison study, both rollers were used along the enterline of the lane. It appears that, for the smaller roller, speial efforts must be made to ensure seating at the edges. In California (J, 2), a vibratory sheepsfoot roller weighing 44, lb was used. Ten rolling passes were applied in eah half of a 12-ft lane. The roller width of 8 ft resulted in 25

4 26 TRANSPORTATION RESEARCH RECORD 1178 TABLE 1 SUMMARY OF ANALYSES OF DEFLECTION MEASUREMENTS: 1--64, JEFFERSON AND SHELBY COUNTIES PARTICLE SIZE (ItUES) * * * lleouical DEA.ECTICJIS STIFFtESS t-m.ll (KS!) MDIETICAL TER-IINI FIELD DEA.ECTI~ ASPHALTIC o~ (I~C~I~5) SlRFACE TEST TEW. DIREC- BEGIN ~:~~= X l0" HZb --~-~-- R:C R:C CRUSIED -~ [ll\te F TI CJ.I M' M' NO.l N0.2 ~.3 N0.4 L[ll\O!NG UWJING l.nlro<en CR<l!X/SEAT STCtE SlllGRJl!l: NO.l N0.2 N0.3 N0.2 12/03/82 12/03/82 12/03/82 7 /20/83 7 /20/ /20/ /31/83 10/31/83 11/01/83 11/01/83 8/01/85 8/01/85 8/01/85 8/01/ ffi o l , , 1,850 1,850 1, ,850 1, 1, * UNBROKEN PAVEt-Em a MmEL 400B RCWJ RATER DY~! C LOIO = 600 l bf STATIC L[ll\O = 1670 l bf 25 HZ FREQLENCY O.~ INCHES /IWLITUlE OF VlBAATIG'l b El.ASTIC STIFFNESS AT 0.5 HZ rn;ql.d{y OF L[ll\O!NG 00 PREVAILING TIWERATillE El.ASTIC STIFFNESS AT 25 HZ FREQLENCY OF L[ll\Q!NG 00 PREVAILING TEM'ERATURE SENSOR POSITICJIS: NO. l 5.25 INCf'ES FIU-1 L[ll\O FEET NO HOES FRrn L[ll\O FEET N ItOES FIU-1 L[ll\O FITT N HOES FRrn L[ll\Q FEET 2, 2,700 2,700 2,700 1,700 2,700 2, 1, 1,700 4, 6, 6, 'lf , overlapping of the middle 4 ft and double rolling for that speifi area. Defletion measurements after seating were typially greater than those before seating. It was onjetured that overworking of the raked areas aused a loosening effet. Kentuky experiene with defletion testing before, during, and after seating is summaried in Figures 2, 3, and 4. It has been onjetured that the initial redution and stabiliation of defletions represent initial seating of the raked onrete pavement. The inrease in defletions to levels greater than those before seating generally supports observations elsewhere. These observations are the subjet for some onern about seating requirements. Failure to ahieve proper seating might result in premature and potentially damaging raks within the asphalti onrete overlay as the result of roking of fragments of portland ement onrete. Conversely, overrolling may ause the existing portland ement onrete pavement to beome unbonded from the temperature reinforement steel or to destroy interlok between individual fragments, or both. There is also onern that traffi will eventually overroll the onrete slab, resulting in premature failure of the asphalti onrete overlay. Pratiality tends to ditate use of heavy rollers and a minimum number of passes as opposed to a greater number of passes of lighter rollers. Use of heavy rollers (50 tons or greater) may overload bridges and be less maneuverable in lose onfines. Lighter rollers generally may require more passes to ahieve effetive seating, but the added maneuverability permits more uniform overage of the pavement. Considering experiene in Kentuky and elsewhere (1, 2, 5-7) and results of defletion measurements, it is reommended that the minimum-sied roller for seating be 35 tons. Multitired pneumati rollers are reommended in plae of two-tired rollers, when possible. At least five passes of a 35- ton pneumati-tired roller are reommended, with a staggered (overlapping) pattern to ensure adequate seating at the edges. Three passes of a 50-ton pneumati-tired roller are also a permissible minimum. It should be emphasied that urrent data do not indiate the equivaleny of the stated overages for eah roller sie. Instead, the stated overages are generally optimum on the basis of minimum number of passes (within the limits of pratial onstrution proedures) for eah roller sie relative to the magnitude of defletion after rolling. Short-Term Performane The oldest in-servie setion of broken and seated portland ement onrete overlaid with asphalti onrete was ompleted in Otober None of the pavement setions has been subjeted to an aumulation of fatigue [18-kip equivalent axle loads (EALs)] neessary for the manifestation of visual surfae distresses. Refletion raking of the asphalti onrete overlay, although not speifially assoiated with strutural deterioration,

5 TABLE 2 SUMMARY OF ANALYSES OF DEFLECTION MEASUREMENTS: 1-71, GALLATIN COUNTY lleoretical OEFLECTHNS STIFFt SS rmu (!<SI} TIEClUICAL TE~NI FIELD DEFLECTl~a A.SPtiAL. TIC COCRETE DEFLECTHNS PARTICLE SlRFACI: SIZE TEST IDI'. DIREC- ~GIN ~~~=-~-~~_.'. -0.5~-25 ~~ ~ ~ CRtm:D -----~:_~~==-~:~~---- (HOES) MTE F" TION W W NO.l N0.2 N0.3 N0.4 LOADING L()!l{)ING lffirol<en CRAO</SEAT STCJE SlBGRADE NO.l N0.2 N0.3 N , 89 4, 89 6, 93 4, D * /17 /82 6/17 /82 6/17 /82 6/17 /82 6/ /82 6/ /fr.? / /82 6/ /83 9/13/8.3 6/Wf!S SOJTH SOJTH SOJTH B'J NORTH SOJTH SOJTH SOJTH SOJTH SOJTH SOJTH SOJTH SOJTH SOJTH SOJTH SOJTH 'Xl SOJTH 'Xl 26.7 NffilH NORTH SOJTH 'Xl 21.6 SOJTH SOJTH 'Xl 27.1 SOJTH 'Xl 27.1 SOJTH SOJTH «l 20.7 SOJTH 'Xl 20.l SOJTH N001H ~lh l , , JOO , 1, 1, 1, , , u~~l<en PA'v9ENT a M'.nL 4<nl ROPD RATER OYtW-l!C LCWJ bf STATIC UWJ = 1670 lbf 25 HZ FRECUt«:Y 0.06 HCfES M'lI1UlE OF VISRAllON b ElASTIC STIFF~ AT 0.5 HZ FRE~ OF L()IJ)llll AUJ R \IAILllll TIM'EPATIR: ElASTIC STIFFl'ESS AT 25 HZ FRE{U/{;Y OF L<WJING AUJ ffievailii>(; Tefff!ATU<E SENSOR POSITICNS: NO. l 5.25 It«:fES FIU-1 LO/ID FEET lloes FIU-1 LCIAO FEET HOES FRG1 LCIAO FEET lt«:fes FRl>1 LCIAO FEET TABLE 3 RATIOS OF DEFLECTIONS: AFTER BREAKING AND BEFORE BREAKING Ratios Partile Sensor Route Termini Diretion Sie Date No. 1 No {)4 - West /20/ {)4 - West 7/20/ {)4 - West 7/20/ {)4 -- West 7/20/ {)4 -- East 10/31/ {)4 - East /01/ South 3--{) 6/ / South / South a I South South South North "No breaking. No. 3 No. 4 Avg

6 28 'TRANSPORTATION RESF.ARCH RECORD 1178 " ~ "' 0"' <= w ~ ~ w"' 0 '.._" 0 ~ ~ "' ~ 164, JEFFERSON AND SHELBY COUNTIES ~ SENSORS , " IJ D.. " AVERAGE DIMENSION OF FRAGMENT (INCHES) FIGURE 1 Comparison of ratios of defletions for 1-64, Jefferson and Shelby ounties, and for 1-71, Gallatin County. :- HIDSLAB 1!1 1!1 35-TON ROLLER -e-e- 50-TON ROLLER BB s BEFORE BREAKING EDGE ICORNERSI l!i 1!1 35-TON ROLLER -e-e- SO-TON ROLLER BB BEFORE BREAKING o., u L&.J -' L&.J :!:~ a: a: L&.J > a; O-t---,-~r--,-----.~-.---,-~..---,------,, y~.,--.-~ BB 0 2 " 6 e NUMBER OF PASSES FIGURE 2 Average defletion versus number of roller passes; midslab tests. g :- THIRD POINTS l!i l!i 35-TON ROLLER -e-e- 50-TON ROLLER BB = BEFORE BREAKING llb ll NUMBER OF PASSES FIGURE 3 Average defletion versus number of roller passes; tests at third points on slab. may be aelerated by the aumulation of axle loads. A total of 451 lane-miles was surveyed to determine the extent and severity of refletive raking. The findings of the survey indiate that less than 7.9 lane-miles (one setion of pavement) o.,.... "' u L&.J...J Cl , /~"'"<!1' ~ BB 0 _, ~ JO NUMBER OF PASSES -e FIGURE 4 Average defletion versus number of roller passes; edge (orner) tests. were observed to have anything more than an oasional rak. Craking in this one setion was observed within 6 months after plaement of the final ourse of the asphalti onrete overlay. Measurements indiated very low levels of defletions relative to other setions, suggesting that the existing onrete pavement was not suffiiently broken. Cores from this setion failed to show any raked and broken onrete. Although none of the data ited are onlusive evidene of improper breaking and seating, the aumulation of evidene suggests that the proess was not suitably ompleted in this setion. Refletive raking in less than 2 perent of the surveyed setions with a sampling rate near 50 perent is evidene of the suess of this onstrution proess in the short term. It is antiipated that long-term performane will be more likely a funtion of fatigue. "Overbreakage" in a few isolated areas has resulted in some loalied pavement failures. Strutural Evaluations Seleted pavement setions have been evaluated by defletion testing at various stages of the onstrution proess. Average defletions for a number of setions for two experimental break-and-seat projets are summaried in Tables 1 12

7 Sharpe et al. 29 and 2. Generally, the data may be grouped into the following ategories: Before raking: all setions After breaking and seating: -3- to 12-in. sied fragments -18- to 24-in. sied fragments -30- to 36-in. sied fragments After overlaying: -3- to 12-in. sied fragments -18- to 24-in. sied fragments -30- to 36-in. sied fragments Data may be evaluated from two perspetives: (a) omparisons of defletions for one setion with those of another setion and (b) mathing of measured defletion basins with theoretially simulated defletions to estimate effetive layer moduli. Ratios of defletions for one stage of onstrution to another may be used to evaluate the effiieny of breaking. Data from Tables 1 and 2 were used to determine suh ratios of defletion. These data are summaried in Table 3 and Figure 1. There are onsiderable differenes in breaking harateristis from projet to projet. For example, average ratios of defletions after breaking to those before breaking are summaried as follows: 1-71, Gallatin County -3- to 12-in. fragments: to 24-in. fragments: 1.02 to to 36-in. fragments: 1.03 to , Jefferson and Shelby ounties -6- to 12-in. fragments: 4.69 to to 24-in. fragments: 2.68 to to 36-in. fragments: 2.41 A more detailed summary of these data is given in Table 3 and Figure 5. Ratios of defletions for after breaking, seating, and overlaying to those before breaking also may be omputed. However, these ratios may be more diffiult to interpret beause of the signifiant impat of temperature on the relative elasti stiffness modulus of asphalti onrete. Suh ratios provide meaningful omparisons only when data for all tests are standardied to some referene temperature for the asphalti onrete overlay. Suh analyses are not presented in this paper. Defletion measurements were used to estimate the effetive stiffness moduli for the various layers of the pavement struture by means of bak-alulation proedures (8). There are numerous approahes that may be used, but generally all are iterative and trial-and-error methods. Bak alulations beome more omplex as additional layers are added to the system. The four-layer system, onsisting of asphalti onrete, broken and seated portland ement onrete, rushed stone, and a semi-infinite layer of ompated subgrade, is not yet subjet to routine bak alulation of effetive layer moduli or effetive layer onditions for the Kentuky Model 400 or Model Road Raters. Efforts are urrently under way, however, to develop and refine suh proedures. Analyses presented herein will desribe only those trial-and-error approahes to bak alulation of effetive layer moduli. Information presented in Tables 1 and 2 illustrates average "' :: u!... 1 "' ::E 100 Ill It: iii "' 10! Cl "' Cll It: "' > LEGEND.!.!! t::.!: H * a a s s -g v VJ v :I64 aoth SUBSCRIPTS.ow k- No.of:ITI COUNTS y - No.of :I64 COUNTS El1 El1 IEFflECTIVE STIFFNESS BROKEN AND SEATED l'cc l'aviement FIGURE 5 Average dimension of fragments versus effetive stiffness moduli for raked and seated portland ement onrete pavements; preliminary design riteria. defletions for several setions of broken and seated pavements from aross Kentuky. Also presented in these tables are simulated defletion basins that approximately math the average defletion basins. These theoretial defletion basins were determined on a trial-and-error basis and do not represent results of a routine proedure for the diret bak alulation of effetive elasti layer moduli. These analyses do illustrate, however, some signifiant trends, as follows: 1. There does not appear to be a unique solution for estimation of effetive layer stiffness moduli (i.e., more than one ombination of layer moduli and layer thiknesses will result in defletion basins losely approximating the measured defletion basin). 2. Effetive moduli may be used to "braket" effetive stiffness moduli for the broken and seated onrete pavement. These ranges may be used to estimate appropriate design moduli, as illustrated in Figure 5. SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS Information presented herein douments the observed performane of rigid pavements that have been reyled in plae in Kentuky by breaking and seating followed by an asphalti onrete overlay. Performane is summaried on the basis of observable or visual onditions as well as defletion testing. A total of 451 lane-miles of pavement were visually surveyed to determine the extent and severity of refletive raking. Extensive refletive raking was observed for only one setion involving less than 8 lane-miles, a "failure" rate of less than 2 perent. It was onjetured on the basis of field observations, defletion measurements, and inspetion of ores that the observed refletive raking may have resulted from improper or inadequate breaking or seating, or both. 2

8 30 Some raking was observed in ontrol setions and transition ones where the existing portland ement onrete pavement was not broken and overlay thiknesses were thinning in transition areas. Refletive raking in those areas was expeted. Defletion measurements were obtained before, during, and after breaking and seating, and after plaement of the asphalti onrete overlay. Empirial analyses of these defletions were used to evaluate the effetiveness of breaking and seating and of the overlay with asphalti onrete. These evaluations involved ratios of defletions after breaking to those before breaking, after overlaying to after breaking, and after paving to before breaking. It has been onluded so far that ratios of defletions for before, during, and after breaking and seating ativities may provide meaningful insights relative to the extent and effetiveness of the breaking, seating, and overlaying proedures. It is reommended that onstrution speifiations inlude a maximum fragment sie observable without the aid of a wetted pavement surfae. For suh speifiations to be more effetive, further efforts are needed to develop orrelations of maximum observable fragment sie for an unwetted slab relative to the maximum fragment sie observable for the same slab broken to an aeptable breaking pattern and viewed with the aid of a wetted surfae. Suh observations should be verified by defletion testing. Additionally, speifiations should inlude aeptable ranges of defletion ratios of after breaking (but before overlaying) to before breaking. Rolling is neessary to stabilie the broken pavement. Rollers as small as 35 tons may be permitted. The minimum number of passes for eah roller should be speified. Tentati,vely, three passes of a 50-ton roller and five passes of a 35- ton roller with a staggered (overlapping) pattern over a 12-ft width appear to be appropriate. These reommendations are based on results of defletion measurements. Three passes of the 50-ton roller will not result in an equivalent level of defletion as will five passes of a 35-ton roller. However, five passes of the 35-ton roller with a staggered pattern should result in more onsistent defletion measurements aross the slab. This may be attributed to the greater maneuverability of the smaller roller and potential to provide more uniform overage of the slab. TRANSPORTATION RESEARCH RECORD 1178 The prinipal objetive of this paper is to summarie Kentuky experiene relating to in-plae reyling of rigid pavements. Analyses and evaluations are ontinuing. Existing data bases are still small and limited. It is essential to ontinue building and maintaining long-term performane data. Proposed speifiation riteria must be verified. Efforts to determine the optimum raking sie should ontinue. Development of a model for the strutural behavior of a broken and seated onrete pavement overlaid with asphalti onrete is neessary for development of a rational thikness design proedure. Proedures for evaluation and bak alulation of the effetive behavior of suh pavements are needed. REFERENCES 1..Craking and Seating of PCC Pavements Prior to Overlaying with Hot Mix Asphalt: State of the Art. Information Series 91, National Asphalt Pavement Assoiation, Riverdale, Md., Craking and Seating of PCC Pavements Prior to Overlaying with Hot Mix Asphalt: State of the Art. Information Series 98, National Asphalt Pavement Assoiation, Riverdale, Md., T. C. Hale. Craking and Seating, Remedy or Rhetori? Constrution Digest, April 4, Speial Provision No. 77A (85), Breaking and Seating of Existing Conrete Pavement. Department of Highways, Kentuky Transportation Cabinet, Frankfort, Ky., Feb R. A. Ekrose and W. E. Poston. Asphalt Overlays on Craked and Seated Conrete Pavements. Donohue and Assoiates, In., Elkhorn, Wis., April E. 0. Lukanen. Strutural Evaluation of Crak and Seat Overlay Pavements. Midwest Pavement Management, In., Sept Crak and Seat Performane. Review Report, Federal Highway Administration, Demonstration Projets Division and Pavements Division, U.S. Department of Transportation, Washington, D.C., April G. W. Sharpe, M. Anderson, R. C. Deen, and H. F. Southgate. Nondestrutive Evaluation of Rigid Pavements Using Road Rater Defletions. In Pro., Third International Conferene on Conrete Pavement Design and Rehabilitation, Purdue University, West Lafayette, Ind., April 23-25, Publiation of this paper sponsored by Committee on Flexible Pavement Constrution and Rehabilitation.